Physicians grappling with uncertainty about artificial intelligence’s role in medicine need look no further than breast radiologist Kathy Schilling, M.D., FACR, for guidance. As medical director of Lynn Women’s Health & Wellness Institute and Lynn Cancer Institute at Boca Raton Regional Hospital, part of Baptist Health, Dr. Schilling has witnessed AI’s transformative potential firsthand.

Her message to colleagues is unequivocal: “Embrace the evolution. Be open, be accepting and actively engage because artificial intelligence is not going away,” she says. For physicians, the question isn’t whether AI will impact their practice, but how quickly they can harness its power to deliver better patient care.

As an early adopter of AI, Dr. Schilling says the technology has made her a better physician. “Every single day, I am grateful that I am working with artificial intelligence to help me in assessing the findings on mammograms,” she says. “Not only are we finding more cancers, but we are finding them earlier and we are decreasing the number of false positives.”

Prior to AI, Dr. Schilling and her team scanned through tens of thousands of black-and-white images a day with digital breast tomosynthesis, viewing 1 mm-thick images through the breast, hunting for a needle in a haystack. Their eyes were strained and fatigued, their concentration on the verge of cracking. Yet they persevered because as breast radiologists, lives depend on their vigilance and precision.

Approximately 40 million screening mammograms are performed in the U.S. each year, according to the U.S. Food and Drug Administration. Mammography remains the gold standard screening technology for breast cancer, which is the second-most deadly cancer in women.

As mammography technology has advanced over the last decade, it has created a tremendous demand on breast radiologists, who are in short supply. In 2014, for example, the move from 2D to 3D mammography increased the number of images for each patient from four to up to 300, Dr. Schilling said. Keeping up with the volume was nearly impossible.

AI flags subtle abnormalities that might otherwise escape even the most experienced eyes and prompts radiologists to investigate further. “When it identifies an abnormality on the mammogram, it will encircle it so that we know exactly where it’s located,” Dr. Schilling says. “Then we take a closer look.”

Dr. Schilling began using AI in 2020 and knew she’d found what she calls her “co-pilot.” But it didn’t occur on Day 1. “It took us maybe several months to be able to trust it and to learn its strengths and weaknesses,” she says. “There are many different uses for AI and it’s only going to become more important in having a great partnership.”

In her work, it’s the partnership between human expertise and machine intelligence that has made a difference in care, she says. In addition to locating a suspicious lesion on a mammogram, AI also categorizes it, giving it a score from 1 to 100.

“The higher the number, the more likely we’re dealing with a malignancy,” she explains. “But it’s up to the radiologist to bring the knowledge that we have about the specific patient. It’s not just black and white. The computer says yes, there’s something or no, there’s not, but the radiologist knows the patient — what her risks are, whether she’s had cancer before, her co-morbidities, her age, if she is frail. We bring the human aspect to care.”

As a physician-scientist, Dr. Schilling always looks for scientific evidence to support the use of new technology. She began researching the Institute’s results while using AI. At the time of the study, the Institute had nine dedicated breast radiologists, with an average of 22 years of experience. With AI, the radiologists:

• Increased cancer detection by 23 percent
• Caught many small cancers — 3 or 4 mm
• Found 33 percent more cancers in patients with dense breasts. Increased breast density makes it more difficult to identify abnormalities
• Discovered 100 percent more invasive lobular cancers, which are difficult to detect because they do not form a typical discrete mass localized in the breast but infiltrate single file through the breast. “Finding these cancers early is important because when they present late, outcomes tend to be poor because they are much larger and more difficult to treat due to delay in diagnosis,” Dr. Schilling says.
• Outcomes were realized without a change in recall rate

“By adding this technology to our toolbox, we are making huge leaps. We are finding cancers years earlier than before. It’s a win-win for all of us,” Dr. Schilling says.

Every woman should have a conversation with her primary care physician or OB/GYN by age 30 about her risk factors for breast cancer. The talk should occur earlier for those with a family history, known genetic mutation or other concern that may predispose them to cancer.

For women at average risk of breast cancer, the current guidelines by the American College of Radiology and the American Society of Breast Surgeons recommend that annual screening begin at age 40. Appointments for screening mammograms are available at Baptist Health Outpatient Services locations at BaptistHealth.net/Mammo or through the PineApp.